// learn.shawon.ch / evidence-desk / the-protocol SYNTHESIS
← Evidence desk

Evidence desk · Personal synthesis

The protocol

The synthesis

Six source cards, collapsed into one protocol. For each lever, the best-supported recommendation — credited to where it actually comes from, and filtered through your Life OS (the cut on Zepbound, ApoB 89, HOMA-IR 3.65, the face-protection floor). Age-old advice is credited to Grandma's wisdom; specific calls go to the person who earned them.

The headline, in one breath: the free fundamentals carry most of the benefit, a single specialist (mostly Attia) tunes the ~20% that matters for your risk, a few things are optional, and a long list of hyped, expensive, or unproven interventions are safe to ignore. Do the boring part first; everything below is sorted by how much it actually moves the needle.

What's inside

  1. How the payoff stacks up
  2. The foundation
  3. Precision for your risk
  4. The supplement stack
  5. Worth considering
  6. Stay away from the hype
  7. The sources, one line each

How the payoff stacks up

The single most useful idea on the whole desk: leverage is wildly uneven. The free foundations do most of the work; the precision layer is a cheap, targeted add-on for your specific risk; the rest is optional or noise. Sorted by real-world payoff, it looks like this — and notice the hype is the thinnest bar.

Foundations — most of the payoff free Precision for your risk cheap, targeted Worth considering optional Skip — hype & unproven
Where the payoff really is: most of it is free; the hype is the sliver at the bottom.

The foundation

Non-negotiable, free, and the highest-evidence part of everything here. Mostly Grandma — every guru on the desk is, at heart, restating these. Do them first; you mostly already do.

Do these first
LeverYour callFrom
Diet baseWhole, minimally-processed food, protein-forward, mostly plants; cut added sugar and ultra-processed food. (Your HOMA-IR makes the sugar discipline the single highest-value food habit.)Grandma
MovementDaily Zone 2 (walking counts) + resistance training 2–4×/week + the occasional all-out sprint.Grandma · Attia · Sisson
SleepEnough, and on a regular schedule — same hours daily. Cheapest high-leverage win if it isn't already locked in.Grandma · Johnson
Smoking & alcoholDon't smoke; keep alcohol low (less is better). The biggest free levers — you're already past them.Grandma
Social connectionDeliberate time with people you love — treat it like a health metric, not an afterthought. The lever the whole desk under-weights.Grandma
Stress & daylightManage stress; get daylight for circadian rhythm — but no deliberate burning (see the skip list, given your tretinoin).Grandma

Precision for your risk

The ~20% the fundamentals can't see — tuned to your panel. This is almost entirely Attia's contribution, because your roadmap already is his framework. Confirm anything touching meds with your prescriber.

Tuned to your numbers
LeverYour callFrom
ApoBTreat ApoB (89, "watch") as the target — not body weight — and drive it toward <70. South-Asian + insulin-resistant is exactly the higher-risk profile where lowering pays; discuss a statin ± ezetimibe with your doctor.Attia
Lp(a)Measured once (15 nmol/L, optimal) — done. No further action.Attia
Insulin resistanceReverse it (HOMA-IR 3.65, TCF7L2): the Zepbound cut, fasted Zone 2, fiber, protein, muscle. Your whole plan already is the right response — keep going.Attia
Protein + muscle on a GLP-1Hold ~170 g protein/day (~2 g/kg) plus progressive resistance training to defend lean mass through the cut — the GLP-1's one real risk.Attia · Sisson
Body compositionTrack visceral fat on DEXA (VAT 65.6 → <52); stop the cut at 14–15% to protect face and hormones — don't chase lower.Attia · your filter
FiberAdd viscous fiber (psyllium) for ApoB/LDL, glycemic load, and GLP-1 regularity. Cheap, multi-purpose.Grandma · the desk
Creatine5 g/day — the highest-confidence supplement for strength and lean mass; especially worth it on a cut.broad consensus
Vitamin DYou're at 59 ng/mL — maintain, don't escalate. VDR reduced-function makes a sensible target reasonable, but dose to the number, not above it.Patrick / Johnson, dialed down
Omega-3Index already optimal — maintain, don't chase a higher number (see the fish-oil card).Patrick · Attia

The supplement stack

Everything above blends behaviours, screening, and pills. Here's just the supplements, pulled into a stack you can actually run — what's worth it for anyone, what's worth it for you, and how it travels across your phases. Every row links to its full verdict on the evidence desk.

Core — run every phase 5 keepers Situational as needed Trim first collagen, HMB Skip — MVI, threonate, GlyNAC, peptides
Four tiers: a small core you run always, a couple of situational adds, two riders to shed after the cut, and a skip pile.
For the general population — start here
SupplementFor most?Only if…
Creatine 3–5 gYesThe one near-universal pick — strength, lean mass, maybe cognition. Cheap, safe, well-replicated.
Protein powderIf short…you can't hit a daily protein target from food. A convenience tool, not magic.
Vitamin DIf low…a blood test shows a real shortfall. Dose to the number, don't megadose.
Omega-3If no fish…you rarely eat oily fish. Food first; the pill fills the gap.
Psyllium fibreIf low-fibre…your diet is short on fibre. Cheap and multi-benefit.
Magnesium (glycinate)If low intake…repletion only; effects are modest.
Multivitamin & most othersNoInsurance you rarely need — test and target a specific gap instead.

The rule underneath the table: food, sleep, and training first. A supplement earns a slot only by clearing the evidence bar (creatine) or filling a measured gap. Most people who eat well need little here.

Your version, re-derived against your numbers (170 g protein, ApoB 89, HOMA-IR 3.65, vitamin D 59, omega-3 already optimal, Phase 1 cut on Zepbound) and sorted into four tiers:

Your stack — tiered
TierSupplementYour call
CoreCreatine 5 gStrongest lever — defends lean mass + strength on the cut. Never cycle off.
Whey / proteinThe tool to hit your 170 g floor against suppressed appetite. The macro the GLP-1 fights you on.
PsylliumOne scoop, three of your levers at once — ApoB/LDL, glycaemic load, GLP-1 regularity.
Vitamin D3Maintain ~59 — dose to the number, don't escalate. K2 optional.
Omega-3Maintain — you're already optimal. Step the double dose down to single and retest.
SituationalMagnesium glycinateOptional repletion; mild rationale on a cut (less food). Add only if sleep or intake warrants.
L-theanineAs-needed calm/focus — with coffee, or to soften the Vyvanse edge. Low stakes.
Trim firstCollagen ~11 gWeak, funding-skewed skin lever — a cheap rider, the first scoop to drop.
HMB 3 gRedundant once protein + training + creatine are running. The other scoop to drop.
SkipMultivitamin · Mg L-threonate · GlyNAC · peptidesNo clear case for you — plus the testosterone boosters and alpha-GPC in the skip list below.

The refinement — what to actually change

You're already running the core well; this is about shedding the dead weight, not adding more. Three concrete moves:

Net: a leaner shake — whey + creatine — with psyllium, vitamin D, and omega-3 around it. Five core items, doing real work, each tied to a number you track.

Carrying the stack across your phases
SupplementCut (now)BuildMaintain
Creatine 5 gkeepkeepkeep
Whey / protein~2–2.4 g/kg~1.6–2 g/kg~1.2–1.6 g/kg
Psylliumkeepkeepkeep
Vitamin D3maintain ~59maintainmaintain
Omega-3 (single)maintainmaintainmaintain
Magnesium glycinateoptionaloptionaloptional
Collagenoptionaldropdrop
HMBoptionaldropdrop

The honest summary: only protein amount really moves with the phase. The core supplements are phase-invariant keepers, and the two riders are the only things to actively shed once the cut ends.

Worth considering

Real but small — preference, not needle-movers. Fine to do if you enjoy them; nothing here earns a place in the core plan.

Optional, low-stakes
LeverYour callFrom
SaunaEnjoy it as recovery if you like the heat — but don't bank on the Finnish-cohort longevity numbers.Patrick
Time-restricted eatingOptional. On Zepbound, appetite suppression already does the calorie work; a tight window adds little.Patrick · Sisson · Johnson
Cold exposureFine "for the mind"; minimal hard metabolic benefit. Both Sisson and Johnson are honest about this.Sisson · Johnson
Foot / minimalist-shoe workLow stakes. Foot-strengthening is real; transition slowly if curious.Sisson
Disciplined measurementTrack the few decision-relevant markers (ApoB, HOMA-IR, VAT) — the good half of Johnson's ethos. Not everything.Johnson

Stay away from the hype

The point of doing all this homework: a clear, sourced list of what to ignore — unproven, ineffective, inefficient, or actively wrong for you. Each is credited to whoever pushes it, so you know the tell when you hear it again.

Skip these
Skip thisWhyPushed by
Rapamycin for longevityAnimal-data bet with real side effects; both Attia and Johnson tried it and backed off.Attia · Johnson
Young-plasma / "blood boy" exchangesNo human benefit — Johnson ran it on himself and quit.Johnson
Offshore gene therapy (follistatin)Unregulated, irreversible, animal-only data. A flat no.Johnson
Big "longevity" stacks — NMN/NR, spermidine, CaAKG, resveratrolNo human hard-outcome data; Attia calls resveratrol "nonsense."Johnson · marketing
Chasing an epigenetic-clock "age-reversal" scoreNoisy surrogate; "31 years younger" is a category error. Track ApoB/HOMA-IR/VAT instead.Johnson
Whole-body MRI & multi-cancer blood testsOverdiagnosis-prone; the Galleri RCT just failed. Not for you at 41 (a colonoscopy by ~45 is the reasonable piece).Attia
"Saturated fat is harmless — eat butter/coconut freely"Settled to raise ApoB; wrong trade for your 89 + South-Asian risk. The Primal idea most worth un-learning.Sisson
"Seed oils are toxic" / avocado-oil necessityLinoleic acid is neutral-to-protective; don't fear-swap to butter (which raises your ApoB). The alarm sells his oil.Sisson
Eliminate grains & legumesNet-protective, high-fiber, high-satiety — keep them, especially on a cut.Sisson
Keto for everyone / "effortless" / endurance boostTies other diets long-term and can spike LDL; you don't need it — Zepbound is doing the calorie work.Sisson · Johnson-adjacent
Testosterone boosters (tongkat, fadogia, zinc-boron)Your T ~597 is fine, and fadogia has zero human safety data.Huberman
Alpha-GPC for focusA TMAO/stroke signal lands badly against your ApoB and CV risk.Huberman
Deliberate sun for vitamin DYou're already replete and on tretinoin (photosensitizing) — a bad trade for you specifically. Ignore "sun prevents skin cancer."Sisson · Patrick
Collagen as muscle protein / "fourth macronutrient"A poor muscle protein; the skin evidence is funding-skewed (see the collagen card).Sisson · Patrick
Peptides (BPC-157 & friends)Anecdote plus preclinical data; never DIY unapproved injectables. Your own rule, and the right one.Reddit hype

How to run it

Foundations are non-negotiable and free — do them first and always. Tune the precision layer with your prescriber against your actual panel (ApoB, HOMA-IR, VAT), not against a podcast. Treat "worth considering" as preference, and the skip list as settled-ignore until something genuinely changes. New claim shows up? Run it through your operating filters and is it strong enough to act? before it earns a slot. The meta-rule from the whole desk: when Grandma and a specialist agree, that's the signal — and when an expensive protocol contradicts her, that's the claim to distrust hardest.

The sources, one line each

The machinery behind this

This page is the output; the engine is your operating filters and the research-methods toolkit — the evidence ladder, surrogate vs hard outcomes, the funding filter, and is it strong enough to act? Re-derive any row yourself.

Built from the six source cards on this desk — Grandma's wisdom, Attia, Patrick, Sisson, Huberman, Bryan Johnson — each of which carries the primary-source citations for its claims. A personal synthesis, not medical advice; confirm anything touching your medications with your prescriber.

Learn · Shawon Chowdhury · a personal protocol, kept rough on purpose · not medical advice